Geometry: Euclid and Beyond by Robin Hartshorne, Springer- Verlag, New York, 2000, Xi+526, $??.??, ISBN 0-387-98650-2
Total Page：16
File Type：pdf, Size：1020Kb
Appeared in Bulletin of the A.M.S., 39 (October 2002), pg 563-571 Geometry: Euclid and Beyond by Robin Hartshorne, Springer- Verlag, New York, 2000, xi+526, $??.??, ISBN 0-387-98650-2 Reviewed by David W. Henderson Introduction The first geometers were men and women who reflected on their experiences while doing such activities as building small shelters and bridges, making pots, weaving cloth, building altars, designing decorations, or gazing into the heavens for portentous signs or navigational aides. Main aspects of geometry emerged from three strands of early human activity that seem to have occurred in most cultures: art/patterns, building structures, and navigation/star gazing. These strands developed more or less independently into varying studies and practices that eventually were woven into what we now call geometry. Art/Patterns: To produce decorations for their weaving, pottery, and other objects, early artists experimented with symmetries and repeating patterns. Later the study of symmetries of patterns led to tilings, group theory, crystallography, finite geometries, and in modern times to security codes and digital picture compactifications. Early artists also explored various methods of representing existing objects, and living things. These explorations led to the study of perspective and then projective geometry and descriptive geometry, and (in 20th Century) to computer-aided graphics, the study of computer vision in robotics, and computer-generated movies (for example, Toy Story). Navigation/star gazing: For astrological, religious, agricultural, and other purposes, ancient humans attempted to understand the movement of heavenly bodies (stars, planets, sun, and moon) in the apparently hemispherical sky. Early humans used the stars and planets as they started navigating over long distances; and they used this understanding to solve problems in navigation and in attempts to understand the shape of the Earth. Ideas of trigonometry apparently were first developed by Babylonians in their studies of the motions of heavenly bodies. Even Euclid wrote an astronomical work, Phaenomena, in which he studied properties of curves on a sphere. Navigation and large-scale surveying developed over the centuries around the world and along with it cartography, trigonometry, spherical geometry, differential geometry, Riemannian manifolds, and thence to many modern spatial theories in physics and cosmology. Building structures: As humans built shelters, altars, bridges, and other structures, they discovered ways to make circles of various radii, and various polygonal/polyhedral structures. In the process they devised systems of measurement and tools for measuring. The (2000-600 BC) Sulbasutram [Sul] contains a geometry handbook for altar builders with proofs of some theorems and a clear general statement of the “Pythagorean” Theorem. Building upon geometric knowledge from Babylonian, Egyptian, and early Greek builders and scholars, Euclid (325-265 BC) wrote his Elements which became the most used mathematics textbook in the world for the next 2300 years and codified what we now call Euclidean geometry. Using Elements as a basis in the period 300 BC to about 1000 AD, Greek and Islamic mathematicians extended its results, refined its postulates, and developed the study of conic sections and geometric algebra, what we now call "algebra". Within Euclidean geometry, there later developed analytic geometry, vector geometry (linear algebra and affine geometry), and algebraic geometry. The Elements also started what became known as the axiomatic method in mathematics. Attempts by mathematicians for 2000 years to prove Euclid's Fifth (Parallel) Postulate as a theorem (based on the other postulates) culminated around 1825 with the discovery of hyperbolic geometry. Further developments with the axiomatic methods in geometry led to the axiomatic theories of the real numbers and analysis and to elliptic geometries and axiomatic projective geometry. The Book The book under review, Geometry: Euclid and Beyond, is situated in this “Building Structures” historical strand of geometry. The author states in his Preface: “In recent years, I have been teaching a junior-senior-level course on the classical geometries. This book has grown out of that teaching experience. I assume only high-school geometry and some abstract algebra. The course begins in Chapter 1 with a critical examination of Euclid's Elements. Students are expected to read concurrently Books I-IV of Euclid's text, which must be obtained separately. The remainder of the book is an exploration of questions that arise naturally from this reading, together with their modem answers. To shore up the foundations [in Chapter 2] we use Hilbert's axioms. The Cartesian plane over a field provides an analytic model of the theory [Chapter 3], and conversely, we see that one can introduce coordinates into an abstract geometry [Chapter 4]. The theory of area [Chapter 5] is analyzed by cutting figures into triangles. The algebra of field extensions [Chapter 6] provides a method for deciding which geometrical constructions are possible. The investigation of the parallel postulate leads to the various non- Euclidean geometries [Chapter 7]. And in [Chapter 8] we provide what is missing from Euclid's treatment of the five Platonic solids in Book XIII of the Elements.” There is almost no mention in this book of the other two strands of geometry. This is a shame since the title (without the subtitle) is Geometry and most of the mathematical research activity in geometry in recent times is situated in the other two strands. A more accurate title would be Companion to Euclid, which happens to be the title of an earlier version of this book that appeared in the Berkeley Mathematics Lecture Notes, volume 9. Euclidean Geometry Until the 20th Century, Euclidean geometry was usually understood to be the study of points, lines, angles, planes, and solids based on the 5 propositions and 5 common notions in Euclid's Elements. In the Elements there is no concept of distance as a real number in the sense we know it today. There is only the concept of congruence of line segments (thus one can say that two segments are equal) and of proportion (so that we can say that two segments are in certain proportion to each other); but we cannot say that they have the same (numerical) length. Geometry as studied in this way is usually called synthetic Euclidean geometry and is the subject of Chapter 1 of Geometry: Euclid and Beyond. Interest in the synthetic geometry of triangles and circles flourished during the late 19th century and early 20th century. One of the best known results of 19th century synthetic geometry is the existence of the nine point circle: Given any triangle in the Euclidean plane, the midpoints of its three sides, the midpoints of the lines joining the orthocenter (the point of intersection of the three altitudes) to its three vertices, and the feet of its three altitudes all lie on the same circle. This nine point circle and similar synthetic Euclidean results are discussed in the last section of Chapter 1. It is usual in schools today for "Euclidean geometry" or just "plane geometry and solid geometry" to not mean synthetic geometry but rather a version of Euclid's geometry with the addition of the real number measure of distances, angles, and areas. This school geometry is a highbred of synthetic (Euclid's) geometry and analytic (or Cartesian) geometry. Though numbers as measure for lengths and areas are not explicit in Euclid’s geometry, they are implicit in his arithmetic of line segments (discussed in Chapter 4) and his “scissors and paste” theory of areas (discussed in Chapter 5). Euclid’s arithmetic of line segments, after defining a unit length determines an ordered field, whose positive elements are the congruence equivalence classes of line segments. Then to develop analytic (or Cartesian) geometry starting from the synthetic Euclidean Plane we choose: 1. A point that we call the origin, O, 2. A segment that we call the unit (length), and 3. Two lines through O, which we call the coordinate axes, (today these are almost always perpendicular, but Descartes did not require them to be so). We can compare a line segment, a, with the unit and define the length of a to be equal to the ratio of a to the unit. Then both coordinate axes can be labeled as number lines with O being the zero. In the usual way we develop the real Cartesian plane and can now study geometric properties using algebra. We can also define a plane geometry over any field by considering its points to be pairs of field elements. This is what Hartshorne does in Chapter 3. In general, different fields give rise to different geometries and these geometries can be used, for example, to study the interdependence of various geometric axioms. The rigorous axiomatic structure that Hartshorne develops in Chapter 2 and further analyzes in Chapter 3 is essentially the axiom system proposed by David Hilbert in 1899 [Hil-a]. It is in this context that Hartshorne defines and discusses the rigid motions: translations, rotations, and reflections. But curiously he does not discuss glide reflections or the classification theorem of rigid motions that states that any rigid motion of the plane is either the identity, a reflection, a translation, a rotation, or a glide reflection. Non-Euclidean Geometries By far the longest chapter in Geometry: Euclid and Beyond is the seventh entitled “Non- Euclidean Geometry”. Notice the use of the singular in his title, as opposed to my title of this section of the review. Wholly within the context of the Building Structures strand it makes sense to talk about the non-Euclidean geometry; but in the other strands there are numerous geometries, that are not Euclidean, such as projective geometry in the Art/Patterns Strand. However, most of the non-Euclidean geometries exist in the Navigation/Star Gazing Strand, as I will discuss: Spherical Geometry Spherical geometry can be said to be the first non-Euclidean geometry.